C1q-Mediated Complement Activation and C3 ... - ACS Publications

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

C1q-Mediated Complement Activation and C3 Opsonization Trigger Recognition of Stealth Poly(2-methyl-2-oxazoline)Coated Silica Nanoparticles by Human Phagocytes. Regina Tavano, Luca Gabrielli, Elisa Lubian, Chiara Fedeli, Silvia Visentin, Patrizia Polverino de Laureto, Giorgio Arrigoni, Alessandra Geffner-Smith, Fangfang Chen, Dmitri Simberg, Giulia Morgese, Edmondo Maria Benetti, Linping Wu, Seyed Moein Moghimi, Fabrizio Mancin, and Emanuele Papini ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.8b01806 • Publication Date (Web): 11 May 2018 Downloaded from http://pubs.acs.org on May 13, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

1

C1q-Mediated Complement Activation and C3 Opsonization Trigger Recognition of Stealth Poly(2methyl-2-oxazoline)-Coated Silica Nanoparticles by Human Phagocytes Regina Tavano1, Luca Gabrielli2, Elisa Lubian2, Chiara Fedeli1,†, Silvia Visentin1, ††, Patrizia Polverino De Laureto3, Giorgio Arrigoni1, Alessandra Geffner-Smith1, Fangfang Chen4,5, Dmitri Simberg4, Giulia Morgese6, Edmondo M. Benetti6, Linping Wu7,8, Seyed Moein Moghimi4,8,9,*, Fabrizio Mancin2, Emanuele Papini1*

1

Department of Biomedical Sciences, University of Padua, Padua, Italy Department of Chemical Sciences, University of Padua, Padua, Italy 3 Department of Pharmaceutical Sciences, University of Padua, Padua, Italy 4 Translational Bio-Nanosciences Laboratory and Colorado Center for Nanomedicine and Nanosafety, The Skaggs School of Pharmacy and Pharmaceutical Sciences, Department of Pharmaceutical Sciences, University of Colorado Denver, Anschutz Medical Campus, 1250 East Mountview Blvd., Aurora, CO 80045, USA 5 Department of Gastrointestinal Surgery, China-Japan Union Hospital, Jilin University, 126 Xiantai Street, Changchun, Jilin, 130033, China 6 Department of Materials, ETH, Zurich, CH 7 Guangzhou Institute of Biomedicine and Health, Chinese Academy of Sciences, Guangzhou, 510530, People's Republic of China 8 School of Pharmacy, Newcastle University, Newcastle upon Tyne NE1 7RU, UK 9 Institute of Cellular Medicine, Newcastle University, Framlington Place, Newcastle upon Tyne NE2 4HH, UK 2

*Correspondence: [email protected] [email protected] †present address: Instut de microbiologie, CHUV, Rue du Bugnon 48, CH-1011 Lausanne, CH ††present address: School of Biological Science, College of Science and Engineering, University of Edinburgh, Edinburgh EH9 3BF, United Kingdom

1 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 38

2 ABSTRACT Poly(2-methyl-2-oxazoline) (PMOXA) is an alternative promising polymer to poly(ethylene glycol), (PEG), for design and engineering of macrophage-evading nanoparticles (NPs). Although PMOXA engineered NPs have shown comparable pharmacokinetics, and in vivo performance to PEGylated stealth NPs in the murine model, its interaction with elements of the human innate immune system has not been studied. From a translational angle, we studied the interaction of fully characterized PMOXA-coated vinyl-triethoxysilane-derived organically modified silica NPs (PMOXA-coated NPs) of approximately 100 nm in diameter with human complement system, blood leukocytes and macrophages, and compared their performance with PEGylated and uncoated NP counterparts. Through detailed immunological and proteomic profiling we show that PMOXA-coated NPs extensively trigger complement activation in human sera exclusively through the classical pathway. Complement activation is initiated by the sensing molecule C1q, where C1q binds with high affinity (Kd = 11 ± 1 nM) to NP surfaces independent of immunoglobulin binding. C1q-mediated complement activation accelerates PMOXA opsonization with the third complement protein (C3) through amplification loop of the alternative pathway. This promoted NP recognition by human blood leukocytes, and monocyte-derived macrophages. The macrophage capture of PMOXA-coated NPs correlates with sera donor variability in complement activation and opsonization, but not with other major corona proteins, including clusterin and a wide range of apolipoproteins. In contrast to these observations, PMOXA-coated NPs poorly activated the murine complement system, and were marginally recognized by mouse macrophages. These studies provide important insights into compatibility of engineered NPs with elements of the human innate immune system for translational steps.

Keywords: C1q, C3, complement, human macrophages, polyoxazoline, stealth polymers

2 ACS Paragon Plus Environment

Page 3 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

3

The interaction of intravenously injected nanoparticles (NPs) with the host blood proteins has long been suggested to modulate NP pharmacokinetics, and their extent of clearance from systemic circulation by macrophages of the reticuloendothelial system (RES).1,2 The rapid clearance of therapeutic NPs by macrophages of the RES, however, could limit their delivery to target sites outside the liver and the spleen.1–4 Steric stabilization of NPs with long-chain hydrophilic polymers/copolymers has become a successful approach in minimizing protein binding, and modulating NP pharmacokinetics and biodistribution. Among many engineered polymers, poly (ethylene glycol), (PEG), has widely been used for design and surface engineering of long-circulating NPs and vesicular drug carriers.5 A classic example is the regulatory-approved Doxil® (a PEGylated liposomal formulation of doxorubicin), and its generic versions.6 Although PEGylation can successfully reduce adsorption of blood proteins to NPs, numerous studies have shown that opsonization of PEGylated NPs by the third protein of the complement system (C3) and its cleavage products (C3b/iC3b) may still occur, and this could lead to their efficient capture by human macrophages.7,8 It is also crucial to consider that some individuals may have antibodies that could recognize the PEG moiety of PEGylated nanopharmaceuticals. This may initiate macrophage recognition of PEGylated NPs through complement opsonization.9–13 Furthermore, PEG can undergo oxidative degradation, and this may increase its complement activation property.14 A number of animal and clinical studies have also indicated that PEGylated NPs may induce cardiopulmonary disturbances, and distress. Earlier studies postulated that complement activation, and subsequent release of C3a and C5a anaphylotoxins induced by PEGylated NPs, may play a causal role in initiating such reactions.7,8,15,16 However, a recent study has demonstrated a transitional link between robust NP clearance from the blood by strategically placed macrophages in vasculature, and adverse hemodynamic reactions independent of complement activation.17 This further indicates that PEGylated NPs may not necessarily behave as stealth entities depending on microenvironmental conditions. In line with the above-mentioned issues, alternative strategies are being considered for design, and engineering of long-circulating and macrophage-evading NPs.18 Among many approaches in NP surface engineering is surface functionalization with polyoxazoline polymers and their derivatives, which has generated promising outcomes.19–27 For instance, polyoxazoline-coated NPs show considerable resistance to protein binding, and are more amenable to further chemical 3 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 38

4

modifications compared with PEGylated NPs.28–30 Furthermore, polyoxazoline-coated NPs exhibit prolonged circulation profiles on intravenous injection in the murine model with comparable pharmacokinetic profiles to PEGylated NPs.27,31,32 Murine models, and murine- and bovine-derived materials are widely used in nanomedicine evaluations. However, there are considerable species differences (e.g., mouse versus human) in innate immune responses to particulate invaders, including complement activation, C3 opsonisation processes, and phagocytic clearance. For instance, although the uptake of superparamagnetic iron oxide nanoworms by both murine and human leukocytes is C3dependent, there are major differences in pathways of complement activation, and the extent of C3 fixation between these species.33 Accordingly, species differences in innate immune system function, and performance can modulate NP pharmacokinetics and responses differently. Therefore, prior to translational and clinical studies, it is necessary to confirm stealth characteristics of engineered nanopharmaceuticals, at least with respect to the human complement system, blood leukocytes, and macrophage responses. Here, we have tested stealth characteristics of poly(2-methyl-2-oxazoline)coated vinyl-triethoxysilane-derived organically modified silica NPs (PMOXA-coated NPs) in human sera from different individuals against complement activation, complement opsonization and dysopsonization processes, and capturing efficacy by human blood leukocytes and monocyte-derived macrophages. Our approach has considered interindividual responses, and highlights important insights into the mechanisms of compatibility of nanomaterials with elements of the human innate immunity, and disparity with the murine system.

4 ACS Paragon Plus Environment

Page 5 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

5

RESULTS AND DISCUSSION Synthesis and Physicochemical Properties of NPs. Polymeric NPs were prepared by ammonia catalyzed microemulsion polymerization of vinyltrietoxysilane (VTES) (Figure 1A, and Figures S1-3).34 Fluorescent labeling and surface functionalization with PEG (MW = 2,000 Da, degree of polymerization = 44), and PMOXA (MW = 4,000 Da, degree of polymerization = 40) were achieved by copolymerizing VTES with Rhodamine B triethoxysilane, and when needed, with the trimethoxysilane derivatives of the polymers. 1H NMR confirmed reaction completion, and the absence of by-products. Broadening of the 1H-NMR polymer signals (with respect to their corresponding line-width in solution), as well as diffusion filter experiments, confirmed NP grafting with designated polymers (Figure S4). All NP types exhibited spheroidal morphology (determined by transmission electron microscopy, Figure S4) with calculated diameter (mean ± SD) of 115 ± 23, 90 ± 10, and 70 ± 6 (n = 400) for uncoated, PEGylated and PMOXA-coated species, respectively. NP hydrodynamic size distribution and concentration were also measured by Nanoparticle Tracking Analysis (NTA). This modality overcomes intrinsic problems observed when DLS is applied to heterogeneous samples, since it is based on video tracking of the Brownian motion of single NPs.35. The NTA results revealed mean hydrodynamic diameters (mean ± SD) of 144 ± 5 (mode 124 ± 8), 117 ± 4 (mode 103 ± 2), and 86 ± 4 (mode 83 ± 2) for uncoated, PEGylated and PMOXA-coated species (n = 3 measurements in all cases), respectively. Size distributions were almost symmetrical, with a ±D50% of 20-25 nm, for all NPs. Dynamic laser light scattering yielded comparable results to NTA for all NP preparations with polydispersity indices 90%) on calcium chelation, and further inhibited on EDTA treatment of serum (Figures 3A, S5 and S6). In contrast to complement proteins, other predominantly deposited species such as clusterin, lipoproteins (LPs), serum albumin and immunoglobulins (Igs) are marginally affected on divalent cation chelation, regardless of NP type (Figure 3A, S5 and S6). MS spectrometry analysis, after in-gel digestions of major electrophoretic bands obtained by separation of the NP-bound HS polypeptides by SDS-PAGE, further agreed with shot-gun analysis (Figure 3B, S7 and S files 4-7). Notably, the bands corresponding to C3 β chain (Mw ~ 65 KDa) and C3c α’ chain fragment 2 (Mw ~ 39 KDa) were only detected under reducing conditions, which confirm C3 cleavage, and covalent association of C3b with NP surfaces. Densitometry analysis of samples incubated in the presence of chelating agents, once again showed that only the binding of complement proteins to PMOXA-coated NPs is strongly reduced on Ca2+ chelation, and further inhibited on EDTA treatment (Figure 3B). The Role of Human Serum and Complement Fixation on PMOXA-coated NP Uptake by Human Phagocytes. Circulating blood phagocytes (monocytes and PMNGs-Polymorphonuclear Granulocytes) and human macrophages showed preferential tendency to capture PMOXA-coated NPs and PEGylated NPs, compared with uncoated NPs, pre-treated with HS (75% v/v) for 15 min at 37°C in 7 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 38

8

the absence of cells, and then incubated with cells after a ten folds dilution in protein free culture medium (Figure 4A&B). Confocal fluorescent microscopy confirmed internalization of all preopsonized NPs, and their localization to the acidic endosomal-lysosomal compartments (Figure 4C). Cell viability was further confirmed on NP uptake studies (Figure S8). Pre-opsonized NP uptake by both blood phagocytes, and human macrophages, however, was dramatically reduced on Ca2+ chelation of serum (Figure 4B). This observation may suggest a likely role for complement activation, and surface C3 fixation, in NP recognition by human phagocytic cells. In contrast to human phagocytes, PMOXA-coated NPs were not only poor activators of the mouse complement system, but were also poorly recognized by mouse monocyte-derived macrophages compared with uncoated NPs (Figure 4D). Next, we investigated whether complement activation and surface C3-fixation play a critical role in NP recognition by human phagocytes. Proteomic studies did not show significant surface association of collectins and associated proteases (e.g., mannose binding lectin (MBL), ficolins, collectin 11 and MBL-associated serine proteases, MASPs)38,39 with NPs. At first instance, this observation excludes the potential involvement of these molecules, and hence the lectin pathway, in complement activation, and C3 fixation. However, on the basis of observed C2 discrepancy between shot-gun proteomics, and functional complement activation studies, we sought to further investigate whether the lectin pathway of complement plays a role in NP recognition by macrophages. We showed no effect of mannose and N-acetyl glucosamine (which compete with the binding of MBL and ficolins to their substrates)8,39 on NP uptake (Figure S9). Furthermore, aprotinin (a MASP inhibitor)40 did not affect the capture of PMOXA-coated NPs by macrophages (Figure S10). Following these studies, we measured macrophage uptake of NPs after incubation with HS depleted of various complement proteins (Figure 5). The results showed that the recognition of polymer-coated NPs by macrophages is blocked by >95% on C3-depletion, thereby suggesting a role for C3b/iC3b as the major opsonic molecule and of cellular complement receptors 3 and 4 (Figure 5A). In addition, depletion from HS of either C1q or C4 (which are required for assembly of the classical pathway C3 convertase) or factor B (which is required for increasing the turnover of the alternative pathway, and the amplification loop) strongly inhibited (77-86 %) capture of PMOXA-coated NPs by macrophages (Figure 5A). Addition of depleted factors to respective sera restored macrophage capabilities in capturing polymer-coated NPs, and comparable to levels seen in intact serum (Figure 5A). These 8 ACS Paragon Plus Environment

Page 9 of 38 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

9

observations confirm the role of complement activation, and C3 surface fixation in NP capture by macrophages. Indeed, the results in Figure 5B confirm the notion that restoration of C1q in C1qdepleted serum triggers complement activation through the classical pathway, and subsequent C3 cleavage. Antibodies of both IgG and IgM classes are known to trigger activation of the classical pathway by facilitating C1q docking. Accordingly, complement activation by PMOXA-coated NPs may have been initiated through binding of either non-specific or putative “PMOXA-specific” antibodies in human serum. Selective depletion of either IgM or IgG from HS did not significantly affect macrophage capture of PMOXA-coated NPs (Figures 5C, and S11). Therefore, C1q-mediated complement activation by PMOXA-coated NPs is apparently antibody-independent. Subsequently, we showed that purified human C1q binds directly to PMOXA-coated NPs, but not to uncoated NPs (Figure S12) with nanomolar affinity (Kd= 11 ± 1 nM; maximal binding = 12 ± 1 C1q molecules/NP). This is rather comparable with the affinity of C1q for some of its natural ligands such as DNA (Kd = 22 nM), histones, Annexins and IgG (Kd range = 1-2 nM)41 and phosphatdylserine (Kd = 37-70 nM)42. For a comparative purpose we also showed low affinity of pooled human IgG and lipoproteins (e.g., HDL) for PMOXA-coated NPs (Table 1, Figure S13). The globular head of C1q is predominantly basic, which makes C1q to function as a charge pattern recognition molecule.43 Indeed, direct binding of the highly cationic C1qA chain to anionic liposomes (e.g., cardiolipin-containing vesicles) was demonstrated earlier, which led to complement activation.44 In addition to this, C1q was also suggested to bind cardiolipin-containing liposomes, and PEGylated vesicles through hydrophobic interactions and hydrogen bonding.44 Given the fact that the zeta potential of PMOXA-coated NPs is close to neutrality, it is most likely that direct C1q association with these NPs could arise from hydrophobic interactions and/or hydrogen bonding by considering that in each PMOXA molecule there are about 40 C=O, and 2 S=O functionalities, and that S=O groups are buried close to NP surfaces, while C=O groups are distributed along the chains. It is also plausible that in serum/plasma, C1q may also function as a charge pattern recognition molecule by binding to exposed anionic domains of protein corona on PMOXA-coated NPs. The Effect of Sera Variability on Complement Activation, Protein Corona and NP Uptake by Macrophages. Next, we compared the effect of sera from 8 individual donors on NP uptake by human phagocytes to account for interindividual variations in complement opsonic activities. The results in 9 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 38

10

Figure 6A show that all 8 sera promoted efficient uptake of PMOXA-coated NPs by macrophages compared with uncoated NPs. In the case of PEGylated NPs, macrophage uptake was more variable, and translated either to smaller, equal or higher values than that of uncoated NPs, depending on serum donor. C3 β chain densitometry analysis of NPs incubated with different sera (Figures 3B, S7 and S14) also mirrored macrophage captures efficacies (Figure 6B). Comparative shot-gun proteomics further confirmed the above mentioned observations, which correlated with C3, and properdin deposition on NPs, and when tested in 3 representative sera displaying different macrophage capturing efficiency [i) a serum that showed low macrophage capture of PEGylated NPs, but high levels

of

PMOXA-coated

NPs

(donor

A;

PEGlow/PMOXAhigh),

ii)

a

serum

promoting

moderate/intermediate capture of PEGylated NPs, yet high levels of PMOXA-coated NPs (donor D; PEG

intermediate

/PMOXA

high

) and iii) a serum promoting high capture of both coated NP types by

macrophages (donor F; PEGhigh/PMOXA high)] (Figure 6C). Furthermore, based on label-free quantification parameter iBAQ, we observed no major quantitative differences in the protein corona for all NP types in the three sera (Figure S15, S files 1-3 and S files 8-12). Subsequently, we compared the polypeptide composition of the corona (Figure S16, S files 1-3 and S files 8-12). Clusterin and lipoproteins constituted major corona proteins from all sera. While clusterin showed no clear preference for any specific NP type, Apo-AIV and Apo-E were consistently enriched on PEGylated NPs, whereas Apo-CI, CII, CIII, Apo-AII were preferentially associated with PMOXA-coated NPs. Uncoated NPs effectively bound all indicated apolipoproteins, but with no clear selectivity. Finally, we statistically correlated the macrophage capture efficacy of polymer-coated NPs, and the amount of those major protein classes consistently found in their corona, through normalization (Figure 6D). As expected, the analysis showed a significant positive correlation between macrophage capture efficacy, and the quantity of deposited C3 (Pearson’s R = 0.9, p